July 21 Update. The July 21, 2017 Congressional Budget Office report did not score the Cruz amendment to the BCRA. The Cruz amendment would allow insurers to sell plans that would not be required to comply with a number of ACA requirements—including guaranteed issue, community rating, preexisting condition coverage, and the essential health benefits—as long as the insurers sold plans that complied with the ACA. It is reported that the CBO would have to take some time to produce a score, in part no doubt because the ACA non-compliant coverage permitted by the Cruz amendment might not qualify as health insurance coverage under CBO definitions.

On July 21, 2017, however, Senators Lee (R-UT) and Johnson (R-WI) circulated to their colleagues a Dear Colleague letter to which they attached an analysis of the Cruz amendment apparently done by the Department of Health and Human Services. (The letter also attaches an analysis of premiums under the ACA done by McKinsey but does not claim that McKinsey is responsible for the Cruz amendment analysis.)

The analysis is remarkable because it shows enrollment growing and premiums falling—even for ACA-compliant plans—under the Cruz amendment. A wide range of observers, including insurers, actuaries, insurance regulators, and consumers have criticized the Cruz amendment as unworkable.

Much of the methodology on which the HHS analysis is based is proprietary and was not disclosed, so it is difficult to verify. The analysis has been criticized, however, on a number of counts. It compares premiums for a 40-year old under the Cruz amendment with average premiums for a 47-year old under the ACA; it incorporates some elements of the ACA and others of the BCRA in its comparisons in a way that distorts the final analysis, and it ignores some problematic features of the BCRA and assuming away problems that would be created by others. The manifold problems with the analysis, which seems to have been written to support the Cruz amendment, rather than to provide impartial analysis, demonstrate again why it is essential to have a nonpartisan, impartial, referee like the CBO to analyze the effects of proposed legislation, a point made in a July 21 letter to Congress signed by all eight former directors of the CBO.

The Latest BCRA Score

On July 20, 2017, the Senate Budget committee released the latest version of the Better Care Reconciliation Act (summary), which was sent to the Congressional Budget Office (CBO) for scoring. It is basically identical to the second BCRA draft posted on July 13, 2017 (which was not separately scored by the CBO) except that it drops the Cruz amendment and includes a couple of small changes in the Medicaid section. The July 13 draft added $70 billion in state long-term stabilization funding to the original BCRA which the Cruz amendment then shifted to the Department of Health and Human Services (HHS) to pay to insurers that offered Affordable Care Act compliant plans. By dropping the Cruz amendment, the July 20 draft leaves these funds with the states, and thus increases state long-term stabilization funding by that amount over the July 13 draft.

Minutes later on July 20, 2017, the CBO released an analysis of the July 20 BCRA version, as did the Joint Committee on Taxation. The biggest change in their analysis from their first BCRA score, which analyzed the original June 22 version, is that the July 20 version would reduce the federal budget deficit by $483 billion over ten years compared to the $321 billion deficit reduction in the original June 22 BCRA.

The biggest reason for the change in the deficit score is that the July 20 version drops provisions of the original BCRA repealing the ACA’s Medicare payroll tax surcharge and unearned income taxes on high wage earners, which would have resulted in the loss of $231 billion in tax revenues. This reduction in lost revenue is partially offset by increases in stabilization funds and funds for opioid treatment. The CBO estimates that $51 billion of the $70 billion in additional stabilization funding would be spent over the ten-year period; the remainder would be spent after 2026. Medicaid cuts remain virtually the same at about $772 billion and net cuts in tax credits other coverage provisions at $408 billion. Federal Medicaid spending would be reduced by 26 percent by 2026, largely because of the phase out of federal Medicaid expansion funding, but also because of the imposition of per capita caps or block grants.

The CBO projections of loss of coverage remain substantially unchanged from the earlier analysis. By 2018, 15 million more people would be uninsured. The number of the uninsured would increase by 19 million by 2020 and 22 million by 2026. By 2026 the percentage of the population that was insured would drop from 90 percent under current law to 82 percent.

Average premiums for benchmark plans (currently 70 percent actuarial value (AV)) would increase by 20 percent by 2018. By 2019 premiums would be only 10 percent higher because the state and insurer stabilization funding would kick in and because the risk pool would be younger because of changes in the tax credits reducing premiums for younger enrollees.

By 2020, premiums would be 30 percent lower than under current law, but they would be purchasing a different product — a 58 percent actuarial value benchmark policy with much higher cost sharing. By 2026 premiums for the higher cost-sharing plan would only be 25 percent lower than under current law as rising health care costs outpaced federal subsidies. Premiums would be somewhat more affordable than they would have been under the earlier BCRA because of the additional state stabilization funding, but the CBO predicts that with lower premiums in the individual market more employers would drop coverage leaving the number of uninsured the same as in the original version.

The effect of the BCRA on premiums compared to current law would vary significantly, however, by age and income. Premiums for a 40 year-old individual with an income at 175 percent of the federal poverty level would drop from $1,700 under current law to $1,450 by 2026, but the premiums would only buy a 58 percent AV plan rather than the 87 percent AV plan to which the individual would be entitled to under current cost-sharing reduction requirements. The 2026 BCRA plan would have a deductible of $13,000 (unless current out-of-pocket limit provisions reduced it to $10,900), about half the individual’s annual income. Under current law, someone at 175 percent of the poverty level would be able to purchase a plan with an $800 deductible. Plans would not be able to offer benefits before the deductible as many do now. The CBO projects that few lower-income people would buy coverage with this little value.

For higher income people net premiums might drop for plans with actuarial values similar to current plans because they would be able to use tax-subsidized health savings accounts to purchase them. The CBO estimates that 75 percent of people able to establish those accounts would use them to pay premiums by 2026. Some employers, however, would drop coverage, increasing taxable income for employees who would no longer receive tax-subsidized coverage.

Premiums would also vary by age, dropping for younger people compared to current law but increasing for older people. For a person with an income at 375 percent of poverty, net premiums would drop for a 21 year-old from $5,100 a year to $3,250, but for a 64 year old, net premiums would increase from $6,750 to $15,300. A 64 year-old not with an income of 450 percent of poverty would see premiums climb from $15,300 to $17,900.

Senate Majority Leader McConnell reportedly intends to bring either the BCRA or the Obamacare Repeal Reconciliation Act to the Senate floor during the week of July 24. The CBO scores are not helpful for either bill.